Automatic driving system and automobile
By adopting an architecture that combines sensor combination with edge nodes in the automobile, edge nodes process data and make decisions, solving the high cost and delay problems caused by the independence of automobile functions, and realizing rapid decision-making and information sharing.
Patent Information
- Application Number
- CN202510258232.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2020-11-04
- Publication Date
- 2025-07-29
AI Technical Summary
The existing automotive functions are independent and data cannot be reused, resulting in high overall costs and centralized processing methods lead to extended business execution time.
The architecture is adopted that combines multiple sensor combinations with edge nodes. The edge node processes sensor data and makes decisions. The central node delegates the decision-making function, and the edge node directly connects to the actuator.
It reduces business execution delay, reduces the overall cost of the car, and realizes information sharing and rapid decision-making among different functions.
Smart Images
Figure CN120382855A_ABST
Abstract
Description
[0001] This application is a divisional application. The application number of the original application is 202011217913.0, and the filing date of the original application is November 4, 2020. The entire content of the original application is incorporated herein by reference. Technical Field
[0002] This application relates to the field of autonomous driving, and particularly to an autonomous driving system and a vehicle. Background Art
[0003] Currently, with the development of the automotive industry, individual vehicles have more and more functions.
[0004] In one solution, the functions of a vehicle develop in a chimney-like manner, that is, different functions are independent of each other. For example, the emergency braking assistance system and the surround view system are independent of each other and have no direct relationship. The data between functions cannot be reused. For example, the perception information determined by the surround view system cannot be utilized by the emergency braking assistance system. For the emergency braking assistance system, a dedicated sensing system for sensing environmental information needs to be set up. As a result, the overall cost of the vehicle is increased.
[0005] In another solution, a centralized processing method is adopted for the functions of a vehicle. That is to say, the decisions of each function are executed by a central node, and the original data collected by various sensors of the vehicle is also processed by this central node. In this solution, the control instructions of the central node are sent layer by layer to the execution components, and the original data collected by the sensors is also reported layer by layer to the central node, resulting in a large service execution delay. Summary of the Invention
[0006] Embodiments of this application provide an autonomous driving system, which can enable the data collected by a sensor to be processed by an edge node directly connected to the sensor to obtain perception information, and the edge node makes relevant decisions based on the perception information, thereby reducing the service execution delay.
[0007] In a first aspect, an embodiment of the present application provides an autonomous driving system, which is configured on a vehicle, and the system includes: multiple sensor combinations, each of the multiple sensor combinations includes at least one sensor; multiple edge nodes, and the edge nodes in the multiple edge nodes correspond to the sensor combinations in the multiple sensor combinations one-to-one; wherein, a first edge node among the multiple edge nodes is used to determine the first perception information of the vehicle based on data collected by a first sensor combination corresponding to the first edge node; a central node connected to the multiple edge nodes, the central node is used to delegate a first state decision function to the first edge node among the multiple edge nodes, so that the first edge node determines that the vehicle enters a first state or the first driving strategy of the vehicle in the first state by executing the first state decision function; wherein, the execution of the first state decision function requires at least the first perception information.
[0008] In a possible implementation, the central node is configured to delegate the first state decision function to the first edge node under the driving state of the vehicle.
[0009] In a possible implementation manner, the central node is configured to determine the first state decision function from a plurality of preset state decision functions according to the first sensor combination.
[0010] In a possible implementation, the system further includes at least one actuator, and the at least one actuator is used to enable the vehicle to enter the first state or to execute the first driving strategy.
[0011] In a possible implementation, the at least one actuator is directly connected to the first edge node; and the central node is configured to determine the first state decision function from a plurality of preset state decision functions according to the at least one actuator.
[0012] In one possible implementation, when the first edge node determines that the vehicle needs to enter the first state or the vehicle needs to execute the first driving strategy, the first edge node sends a control instruction to the at least one actuator, causing the vehicle to enter the first state or causing the at least one actuator to execute the first driving strategy.
[0013] In one possible implementation, the central node is also used to send a second state decision function to a second edge node among the multiple edge nodes, so that the second edge node determines that the vehicle enters the second state or the second driving strategy of the vehicle in the second state by executing the second state decision function; wherein, the execution of the second state decision function requires at least second perception information, and the second perception information is determined by the second edge node based on the data collected by the sensor combination corresponding to the second edge node; the second state is different from the first state.
[0014] In one possible implementation, the execution of the first state decision function also requires third perception information; the third perception information is determined by a third edge node among the multiple edge nodes based on data collected by the sensor combination corresponding to the third edge node; the third edge node is used to send the third perception information to the first edge node when the third perception information is determined.
[0015] In a possible implementation, the first state is any one of the following:
[0016] Over the air (OTA) status, highway pilot (HWP) status, and autonomous emergency braking (AEB) status.
[0017] In a possible implementation manner, each edge node among the multiple edge nodes is directly connected to a sensor in a corresponding sensor combination.
[0018] In a possible implementation, the edge node is an electronic control unit ECU, and the central node is a domain controller.
[0019] In a second aspect, an embodiment of the present application provides an autonomous driving system, which is configured on a vehicle, and the system includes: at least one actuator, used to cause the vehicle to enter a first state or to execute a driving strategy in the first state; a first edge node connected to the at least one actuator; a central node connected to the first edge node, used to delegate a first state decision-making function to the first edge node, so that the first edge node determines whether the vehicle enters the first state or the driving strategy of the vehicle in the first state; when the vehicle needs to enter the first state or the vehicle needs to execute the first driving strategy in the first state, the first edge node sends a control instruction to at least one actuator, so that the vehicle enters the first state or the at least one actuator executes the first driving strategy.
[0020] In a possible implementation, the system further includes: at least one sensor; a second edge node directly connected to the at least one sensor, configured to determine first perception information of the vehicle according to data collected by the at least one sensor, and send the first perception information to the first edge node, so that the first edge node performs the first state decision function based on the first perception information.
[0021] In a possible implementation, the first edge node is directly connected to at least one sensor; the first edge node is further configured to determine first perception information of the vehicle according to data collected by the at least one sensor, and perform the first state decision function based on the first perception information.
[0022] In a possible implementation, the central node is configured to determine the first state decision function from a plurality of preset state decision functions according to the at least one sensor.
[0023] In a possible implementation, the at least one actuator is directly connected to the first edge node; the central node is configured to determine the first state decision function from a plurality of preset state decision functions according to the at least one actuator.
[0024] In a possible implementation, the first edge node is an electronic control unit (ECU), and the central node is a domain controller.
[0025] In a third aspect, an embodiment of the present application provides a self-driving vehicle, including the system described in the first aspect or the system described in the second aspect.
[0026] In the self-driving system provided by the embodiment of the present application, data collected by vehicle sensors can be processed by the edge node closest to or directly connected to the sensors, and relevant decisions are made by the edge node, thereby reducing the latency of the vehicle to execute corresponding services. Alternatively, the edge node directly connected to the actuator can make relevant decisions and hand them over to the actuator for execution, thereby reducing the latency of the vehicle to execute corresponding services. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1A Shows a star interconnection architecture;
[0028] Figure 1B Shows a full interconnection architecture;
[0029] Figure 1C Shows a domain-based interconnection architecture;
[0030] Figure 1D Shows a ring interconnection architecture;
[0031] Figure 2 Shows an autonomous driving system provided by an embodiment of the present application;
[0032] Figure 3 Shows another autonomous driving system provided by an embodiment of the present application;
[0033] Figure 4 Shows yet another autonomous driving system provided by an embodiment of the present application;
[0034] Figure 5 Shows yet another autonomous driving system provided by an embodiment of the present application. Detailed implementation manners
[0035] Next, the technical solutions in the embodiments of the present invention will be described in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all of the embodiments.
[0036] In the description of this specification, "an embodiment" or "some embodiments" etc. mean that in one or more embodiments of this specification, specific features, structures or characteristics described in conjunction with the embodiment are included. Thus, statements such as "in an embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments" etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways.
[0037] Among them, in the description of this specification, unless otherwise specified, " / " means "or", for example, A / B can mean A or B; "and / or" herein is only a description of the association relationship of associated objects, indicating that three relationships can exist, for example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone these three situations. In addition, in the description of the embodiments of this specification, "a plurality of" means two or more than two.
[0038] In the description of this specification, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. The terms "include", "comprise", "have" and their variants all mean "include but not limited to", unless otherwise specifically emphasized in other ways.
[0039] The electrical / electronic architecture (EEA) of an automobile can have a star interconnect architecture, a full interconnect architecture, a domain interconnect architecture, a ring interconnect architecture.
[0040] The star - interconnected architecture can be as Figure 1A shown. It can have one or more central nodes such as central node 110, and can also have multiple edge nodes such as edge node 120, edge node 130, edge node 140, edge node 150, etc. Among them, each edge node can be directly connected to one or more sensors and / or directly connected to one or more actuators. In the star - interconnected architecture, the central node can be directly connected to the edge nodes.
[0041] One or more central nodes such as central node 110, and multiple edge nodes such as edge node 120, edge node 130, edge node 140, edge node 150, etc. can also adopt a fully - interconnected architecture, specifically as Figure 1B shown. Among them, between the central node and the edge nodes, and between the edge nodes and the edge nodes can be connected through an interconnected bus.
[0042] One or more central nodes such as central node 110, and multiple edge nodes such as edge node 120, edge node 130, edge node 140, edge node 150, etc. can also adopt a domain - interconnected architecture, specifically as Figure 1C shown. In this domain - interconnected architecture, central node 110 can be divided into central node 110A and central node 110B. Among them, central node 110A can be used as a domain controller and form a domain with edge nodes such as edge node 120 and edge node 130. Central node 110B can be used as a domain controller and form another domain with edge nodes such as edge node 140 and edge node 150. Nodes under different domains are indirectly connected through the corresponding domain controllers.
[0043] One or more central nodes such as central node 110, and multiple edge nodes such as edge node 120, edge node 130, edge node 140, edge node 150, etc. can also adopt a ring - interconnected architecture, specifically as Figure 1D shown. Different edge nodes can be directly connected, so that a ring network can be formed among multiple edge nodes. Each edge node can be directly connected to the one or more central nodes.
[0044] It should be noted that the connections between the above - mentioned components refer to communication connections. The components can specifically achieve communication connections through a local interconnect network (LIN) bus, a controller area network (CAN) bus, or hard - wiring, etc.
[0045] In the embodiments of the present application, the direct connection between two components means that there is a direct communication link between the two components without the need for other components to forward or transmit information.
[0046] In some embodiments, one or more sensors directly connected to the edge node may include sensors for collecting information about the vehicle's surrounding environment. The information about the vehicle's surrounding environment may refer to information about other vehicles, pedestrians, lane lines, road signs, traffic lights, obstacles, road edges, and other objects in the vehicle's surrounding environment. The information about an object may include the object's orientation and distance relative to the vehicle, and may also include the object's speed and direction of movement, etc., which will not be listed one by one here. The sensors for collecting information about the vehicle's surrounding environment may be cameras, radars, etc. Among them, the radar may be a lidar, a millimeter-wave radar, an ultrasonic radar, etc.
[0047] In some embodiments, one or more sensors directly connected to the edge node may include sensors for collecting information about the vehicle itself. The information about the vehicle itself may include information such as the vehicle's location, driving speed, and driving direction that can be controlled or changed by the vehicle. The information about the vehicle itself may also include information about the vehicle's internal systems, such as the temperature in the passenger compartment, the fuel level in the fuel tank, and the motor temperature. Correspondingly, the sensors for collecting information about the vehicle itself may include positioning components (the positioning component may be a global positioning system (GPS) module, or a Beidou navigation satellite system (BDS), or a Galileo satellite navigation system, etc.), an inertial measurement unit (IMU), a thermometer in the passenger compartment, a fuel gauge, a thermometer at the motor, etc.
[0048] In the embodiments of the present application, one or more sensors directly connected to the edge node may be referred to as the sensor combination corresponding to the edge node. The sensors included in the sensor combinations corresponding to different edge nodes may be different, partially the same, or all the same. The embodiments of the present application do not limit this. In specific implementation, the sensors in the sensor combinations corresponding to each edge node may be set.
[0049] One or more actuators directly connected to the edge node may refer to components that respond to control instructions to perform operations such as decelerating, steering, accelerating, and turning on the air conditioner on the vehicle. Specifically, the actuator may include a braking component, an accelerator, a steering component, an air conditioner switch, etc.
[0050] In the embodiments of the present application, one or more actuators directly connected to an edge node can be referred to as the actuator combination corresponding to the edge node. The actuators included in the actuator combinations corresponding to different edge nodes can be different, partially the same, or completely the same. The embodiments of the present application do not limit this. In specific implementation, the actuators in the sensor combination corresponding to each edge node can be set.
[0051] Based on the above architecture, the embodiments of the present application provide an autonomous driving system. In this autonomous driving system, an edge node can process the data collected by its corresponding sensor combination to obtain the perception information of the vehicle where the autonomous driving system is located; and the edge node can receive the decision-making function delegated by the central node and execute the decision-making function according to the perception information of the vehicle, thereby generating a control instruction for the vehicle to instruct the relevant actuators to perform relevant operations. In this driving system, the data collected by the vehicle sensors can be processed by the edge node closest to or directly connected to the sensors, and relevant decisions are made by this edge node, thereby reducing the latency of the vehicle to execute corresponding services.
[0052] Next, in different embodiments, the autonomous driving system provided by the embodiments of the present application will be introduced by way of example.
[0053] In an embodiment of the present application, the edge node may be a component with data processing capabilities, such as an electronic control unit (ECU). The raw data collected by the sensor combination corresponding to the edge node can be sent to the edge node for processing to obtain the perception information of the vehicle. It can be understood that the raw data collected by the sensor combination may include environmental data of the vehicle's environment and / or information of the vehicle itself. The specific information included in the raw data is determined by the sensor corresponding to the raw data. The sensors connected to the edge node can be specifically referred to the above introduction and will not be elaborated here. The raw data collected by the sensor or the sensor combination can be used to determine the perception information of the vehicle. Exemplarily, it can be set that the sensor combination corresponding to the edge node 120 includes one or more cameras. The environmental image data collected by the one or more cameras can be sent to the edge node 120. The edge node 120 can use computer vision algorithms (such as object recognition algorithms, structure from motion (SFM), etc.) to identify the objects in the environmental image data to obtain the perception information of the vehicle. Exemplarily, it can be set that the sensor combination corresponding to the edge node 120 includes multiple cameras and one or more radars. The environmental image data collected by the multiple cameras and the point cloud data collected by the one or more radars can be sent to the edge node 120. The edge node 120 can use multi-sensor fusion algorithms (such as Kalman filtering algorithms, multi-Bayesian estimation algorithms, neural networks) to process the environmental image data collected by the multiple cameras and the point cloud data collected by the one or more radars to obtain the perception information of the vehicle. In this embodiment, the edge node closest to or directly connected to the sensor or the sensor combination can determine the perception information of the vehicle according to the raw data collected by the sensor or the sensor combination. Thus, the rapid determination of the perception information can be realized, which can or contribute to reducing the service execution delay.
[0054] In some embodiments, when or after determining the perception information of the vehicle, the edge node may send the perception information to other edge nodes so that the other edge nodes can reuse the perception information. Specifically, taking edge node 120 and edge node 150 as examples. Edge node 120 may send the perception information it determines to edge node 150. Thus, edge node 150 can directly make relevant decisions based on the perception information to perform relevant control on the vehicle. Exemplarily, edge node 150 and edge node 120 may pre-establish an information subscription relationship. That is to say, edge node 150 may subscribe to the perception information determined by edge node 120, so that whenever edge node 120 determines the perception information, it can send the perception information to edge node 150. In this embodiment, after determining the perception information of the vehicle, the edge node may share the perception information with other edge nodes, thereby realizing information sharing between different business systems or domains of the vehicle.
[0055] The central node may be a component with data processing capabilities, such as a domain controller. Among them, the central node may have the ability to execute various business functions of the vehicle and also have the ability to delegate one or more business functions to one or more edge nodes. That is to say, in some embodiments, the central node may execute a certain business function by itself. In other embodiments, the central node may also delegate the business function to one or more edge nodes for execution.
[0056] In the embodiments of the present application, the business function may also be simply referred to as a function. In some embodiments, the business function may refer to the function of determining whether the vehicle enters a certain state according to relevant information. In some embodiments, the business function may refer to the function of making decisions or determining the driving strategy of the vehicle in a certain state according to relevant information. In the embodiments of the present application, the business function may also be referred to as a state decision function. Delegating the business function may also be referred to as delegating the state decision function. Delegating the business function or delegating the state decision function may refer to delegating the decision-making authority and decision-making function of the business function. The state may also refer to the driving scenario, which may specifically be a highway pilot state, an over-the-air download state, or an automatic emergency braking state, etc., and will not be listed one by one here. The driving strategy may refer to the specific control of the vehicle during driving, such as accelerating, decelerating, steering, etc., and will not be listed one by one here.
[0057] The downlink operation may consist of one or more operations. In some embodiments, the downlink operation may include the central node sending the code for implementing one or more service functions to one or more edge nodes and sending the running instructions of the code to the one or more edge nodes. Thus, the one or more edge nodes can, in response to the running instructions, run the code to execute the one or more service functions. In some embodiments, the one or more edge nodes may pre-store the code for implementing some or more service functions. The downlink operation may be the central node sending the running instructions of the code to the one or more edge nodes to indicate that the one or more edge nodes can run the code and execute the one or more service functions.
[0058] It can be understood that the purposes of using different sensors may be different. In other words, the service functions (or state decision functions) corresponding to different sensors may be different. Thus, the central node can determine one or more service functions according to the sensors corresponding to the edge nodes and downlink the one or more service functions (or state decision functions) to the edge nodes. Taking the highway pilot (HWP) service function as an example, the corresponding sensor combination A1 includes 3 forward cameras, 4 side cameras, 1 rear camera, 4 surround cameras, 4 lidars, 1 millimeter-wave radar, 16 ultrasonic radars, and 1 inertial measurement unit. It can be set that the sensor combination corresponding to the edge node 120 is the sensor combination A1. The central node 110 can determine that the service function corresponding to the sensor A1 may be the highway pilot service function (or highway pilot state decision function) according to the sensors included in the sensor A1. Thus, the central node 110 can downlink the highway pilot service function (or highway pilot state decision function) to the edge node 120. It should be noted that in the embodiments of the present application, "forward" refers to the front direction of the vehicle, "rear" refers to the rear direction of the vehicle, and "side" refers to the side direction of the vehicle.
[0059] It should be noted that the above is only an example illustration of the correspondence between sensors and service functions, and is not a limitation. In specific implementation, the developer or designer of the vehicle can establish the correspondence between sensors and service functions (or state decision functions) according to experience or experiments.
[0060] In other examples, the central node can also determine the edge node based on the sensors corresponding to over the air (OTA) downloads and delegate the OTA service function (or the OTA status decision function) to the determined edge node. The central node can also determine the edge node based on the sensors corresponding to autonomous emergency braking (AEB) and delegate the AEB service function (or the AEB status decision function) to the determined edge node.
[0061] It can be understood that different actuators correspond to or execute different business functions (or state decision functions). In other words, the actuators used when the vehicle enters different states may be different, and the actuators used by the vehicle to execute driving strategies in different states may be different. For example, actuators such as the throttle and steering components can perform acceleration, deceleration, steering and other operations while the vehicle is in the highway navigation state. In other words, the throttle, steering components, etc. usually correspond to the highway navigation business function (or the highway navigation state decision function). For another example, actuators such as the brake components can perform braking operations while the vehicle is in the automatic emergency braking state. In other words, components such as brakes usually correspond to the automatic emergency braking business function (or the automatic emergency braking state decision function). It should be noted that the above only provides examples of the business functions corresponding to different actuators, and is not limiting. In specific implementation, the developer or designer of the car can establish a correspondence between the actuator and the business function based on experience or experiments.
[0062] In some embodiments, a central node can determine one or more service functions (or state decision functions) based on the actuators connected to an edge node and delegate these one or more service functions (or state decision functions) to the edge node. In this embodiment, the edge node to which the actuator is closest or directly connected can make decisions about the service functions (or state decision functions) and generate control instructions, allowing the actuator to quickly obtain the control instructions and perform operations accordingly, thereby reducing service execution latency.
[0063] In one example of these embodiments, while the vehicle is driving, the central node can determine one or more business functions (or state decision functions) based on the actuator connected to the edge node, and delegate the one or more business functions (or state decision functions) to the edge node.
[0064] The edge node can execute one or more service functions received from the central node. In other words, after the central node issues one or more service functions to the edge node, the edge node can execute the one or more service functions. Exemplarily, when executing the one or more service functions, the edge node can use the sensing information determined by itself for decision-making. Exemplarily, when executing the one or more service functions, the edge node can receive the sensing information determined by other edge nodes and use the sensing information for decision-making. Exemplarily, when executing the one or more service functions, the edge node can receive the sensing information determined by other edge nodes and use the sensing information determined by other edge nodes and the sensing information determined by itself for decision-making.
[0065] In some embodiments, to accelerate the response speed of the edge node to the outside world, the intelligence level of the edge node can be reduced so that the edge node can execute service functions in a simple and direct data processing manner. Thereby, the execution delay of the service function can be reduced. Exemplarily, compared with the edge node, the central node has a higher intelligence level and more powerful data processing capabilities and can be used to execute more complex functions such as prediction and reasoning. Exemplarily, the edge node can be called a low-level control node, and the central node can be called a high-level control node.
[0066] Next, in specific embodiments, the functions of the edge node and the central node are introduced by way of example.
[0067] Refer to Figure 2 , in some embodiments, the autonomous driving system of a vehicle may include a central node 210, an edge node 220, a sensor combination A1, and an actuator combination B1.
[0068] The sensor combination B1 may include one or more sensors. The one or more sensors can be directly connected to the edge node 220 and send the collected raw data A11 to the edge node 220. The edge node 220 may include a sensing information determination module. The sensing information determination module can determine sensing information according to the raw data A11. For details, reference can be made to the above introduction and will not be elaborated here.
[0069] The actuator combination B1 may include one or more actuators. The one or more actuators can be directly connected to the edge node 220. The one or more actuators can respond to control instructions to make the vehicle enter state C1 or execute the driving strategy of the vehicle in state C1. The state may also refer to a driving scenario, which may specifically be a highway pilot state, an over-the-air download state, or an automatic emergency braking state, etc., and will not be listed one by one here. The driving strategy may refer to the specific control of the vehicle during driving, such as acceleration, deceleration, steering, etc., and will not be listed one by one here.
[0070] The central node 210 may include a status decision function management module, which can be used to manage a plurality of preset status decision functions. The central node 210 may determine the status decision function C1 from the plurality of preset status decision functions. The status decision function C1 is used to decide whether the vehicle enters the status C1 and / or the driving strategy of the vehicle in the status C1.
[0071] In some embodiments, the central node 210 may determine the status C1 decision function from the plurality of preset status decision functions according to the sensors in the sensor combination A1. For details, reference may be made to the above description and will not be elaborated here.
[0072] In some embodiments, the central node 210 may determine the status C1 decision function from the plurality of preset status decision functions according to the actuators in the actuator combination B1.
[0073] In some embodiments, when or after determining the status C1 decision function, the central node 210 may send the status C1 decision function to the edge node 220. For details, reference may be made to the above description and will not be elaborated here. Exemplarily, when the vehicle is running, the central node 210 sends the status C1 decision function to the edge node 220.
[0074] In some embodiments, as described above, the edge node 220 may pre-store the code for implementing the status C1 decision function. When or after determining the C1 decision function, the central node 210 may send an execution instruction to the edge node 220 to instruct the edge node 220 to start running the code for implementing the status C1 decision function. For details, reference may be made to the above description and will not be elaborated here. Exemplarily, when the vehicle is running, the central node 210 sends an execution instruction to the edge node 220.
[0075] The edge node 220 may include a decision module. The decision module may execute the status C1 decision function. Specifically, the decision module runs the code for implementing the status C1 decision function. As Figure 2 shown, the decision module may obtain the perception information from the perception information determination module, and based on the perception information, by executing the status C1 decision function, decide whether the vehicle enters the status C1, or decide the driving strategy of the vehicle in the status C1. When or after determining that the vehicle needs to enter the status C1 or the driving strategy of the vehicle in the status C1, the decision module may generate a control instruction C11 for instructing the vehicle to enter the status C1 or for instructing the actuator to execute the driving decision, and send the control instruction C11 to the actuator combination B1. The actuator combination B1 may cause the vehicle to enter the status C1 or execute the driving strategy according to the control instruction C11.
[0076] In an illustrative example, the state C1 is taken as the highway pilot state. The decision-making module, by executing the state C1 decision-making function, determines whether the environment where the vehicle is located is suitable for enabling the highway pilot function, that is, whether it is suitable to enter the highway pilot state, according to the perception information. If the environment where the vehicle is located is suitable for enabling the highway pilot function, the decision-making module can generate a control instruction C11 for instructing the vehicle to enter the highway pilot state and send the control instruction C11 to the actuator combination B1. The actuator combination B1 can enable the highway pilot function according to the control instruction C11.
[0077] In an illustrative example, still taking the state C1 as the highway pilot state. When the vehicle is already in the highway pilot state, the decision-making module can generally execute the state C1 decision-making function and determine whether the vehicle should turn according to the perception information. If the determination result is that the vehicle should turn, the decision-making module can generate a control instruction C11 for instructing the vehicle to turn and send the control instruction C11 to the actuator combination B1. The steering component in the actuator combination B1 can perform vehicle steering according to the control instruction C11.
[0078] It should be noted that the vehicle's autonomous driving system does not only include the edge node 220, but may also include other edge nodes. In some embodiments, the edge node 220 can also send the perception information determined by the perception information determination module to other edge nodes in the autonomous driving system. These other edge nodes can utilize this perception information to execute relevant service functions and make decisions. Among them, the service function can be received by these other edge nodes from the central node 110. Or rather, the central node 110 issues this service function to these other edge nodes.
[0079] The above only takes the edge node 220 as an example to illustrate the functions of the edge node and the relationship between the edge node and the central node. Other edge nodes included in the autonomous driving system can be implemented with reference to the above introduction of the edge node 220, and will not be elaborated here one by one.
[0080] In Figure 2 In the illustrated embodiment, rapid determination of perception information can be achieved, which can reduce or help reduce the service execution delay; and the actuator can obtain the control instruction as soon as possible and operate according to the control instruction. Thus, the service execution delay can be further reduced.
[0081] Refer to Figure 3 In some embodiments, the vehicle's autonomous driving system can include a central node 310, an edge node 320, a sensor combination A1, an edge node 330, and an actuator combination B1.
[0082] The sensor combination B1 may include one or more sensors. The one or more sensors may be directly connected to the edge node 220 and send the collected raw data A11 to the edge node 320. The edge node 320 may include a sensing information determination module. The sensing information determination module may determine the sensing information according to the raw data A11. When or after determining the sensing information, the edge node 320 may send the sensing information to the edge node 330.
[0083] The actuator combination B1 may include one or more actuators. The one or more actuators may be directly connected to the edge node 320. The one or more actuators may respond to the control instruction to make the vehicle enter the state C1 or execute the driving strategy of the vehicle in the state C1.
[0084] The central node 310 may include a state decision function management module, which may be used to manage a plurality of preset state decision functions. The central node 310 may determine the state decision function C1 from the plurality of preset state decision functions. Exemplarily, the central node 310 may determine the state C1 decision function from the plurality of preset state decision functions according to the actuators in the actuator combination B1.
[0085] In some embodiments, when or after determining the state C1 decision function, the central node 310 may send the state C1 decision function or the execution instruction to the edge node 330. Specifically, reference may be made to the introduction of the Figure 2 illustrated embodiment above, which will not be elaborated here.
[0086] The edge node 330 may include a decision module. The decision module may generate a control instruction C11 according to the sensing information received by the edge node 320 by executing the state C1 decision function. The edge node 330 may send the control instruction C11 to the actuator combination B1 so that the actuators in the actuator combination B1 perform related operations according to the control instruction C11. Specifically, reference may be made to the introduction of the Figure 2 illustrated embodiment above, which will not be elaborated here.
[0087] It should be noted that the vehicle's autonomous driving system does not only include the edge node 320 and the edge node 330, but may also include other edge nodes. The above only takes the edge node 320 and the edge node 330 as examples to illustrate the functions of the edge nodes and the relationship between the edge nodes and the central node. Other edge nodes included in the autonomous driving system may be implemented with reference to the introduction of the edge node 320 or the edge node 330 above, which will not be elaborated one by one here.
[0088] In Figure 3 the illustrated embodiment, the actuator can obtain the control instruction as soon as possible and operate according to the control instruction. Thus, the delay of service execution can be reduced.
[0089] Next, in combination with Figure 4 , in a specific example, the autonomous driving system provided by the embodiments of the present application will be illustrated by way of example.
[0090] In this example, the central node and the edge node can be different as follows.
[0091] a1. Different levels of processing intelligence. The nodes are divided into low-level intelligent processing and high-level intelligent processing.
[0092] b1. Different response times: Generally, there are different time requirements for responding to the outside world. Generally, the nodes with faster responses have simple processing functions and low levels of intelligence; for high-level intelligent functions such as prediction and reasoning, the logic is complex and the response is relatively slower.
[0093] c1. Different scenarios: In different scenarios, the purposes of using each sensor are different; the high-level control node sends control commands and local planning and other information to each low-level processing node based on the scenario; it is ensured that key action decisions can be directly made at the low-level processing node.
[0094] In this example, a hierarchical closed loop can be implemented as follows.
[0095] a2. Fast closed loop based on spatial position: Information collection (camera, radar) -> low-intelligence processing -> execution (action) (adjust direction, distance focus on key targets, etc.); the closed loop is completed on the edge node, and the nodes are distributed.
[0096] b2. Global comprehensive closed loop: Information collection (all information collection points, all types) -> high-level intelligent processing -> execution (action) (vehicle control such as acceleration, braking, steering, etc.); the closed loop is jointly completed by the edge and central nodes, and the high-level intelligent processing is centrally deployed.
[0097] c2. Centralized control: Each control node separately sends control information to the vehicle control component, and ensures reasonable output through the vehicle control arbitration mechanism.
[0098] In Figure 4 , the safety system is equal to the low-level control node and is the last node for final control distribution in the system. It can be placed inside the low-level or high-level node, or independently.
[0099] Next, the principle of Figure 4 will be described.
[0100] In the system, relevant sensors or functions are placed at low-level control nodes, and the high-level control node controls the overall system; the overall scene or state management of the system is placed here. After the system is started, the scene information of the system is sent to the corresponding low-level control nodes according to business needs (as shown by the indication line segments 1 and 6 in the figure); each level of nodes in the system starts different functions according to the corresponding scene (as shown by the indication line segment 2 in the figure). For the nodes controlled by the vehicle control, which vehicle control is used as the final control information is confirmed according to the state of the system central node and the state of the corresponding low-level node.
[0101] In the system design stage, according to principles such as business characteristics, resource proximity, resource consumption, and functional safety, the deployment locations of the extended services in the figure are clarified. For example, Figure 4 in it, normal autonomous driving services are as shown by the indication line segments 3 and 5; but at the same time, a set of other functions need to be implemented, and the corresponding process is shown by the indication line segment 4, which may exist simultaneously or independently; this process reduces the business processing process and ensures its own independence.
[0102] Next, combined with Figure 5 In another specific example, taking the autonomous emergency braking (AEB) function as an example, the autonomous driving system provided by the embodiments of the present application will be illustrated.
[0103] Figure 5 This is an application example of the coexistence of an autonomous driving system and assisted driving under a certain network configuration provided by the embodiments of the present application. Among them, the AEB function is deployed to the same edge node or edge safety system as the required sensors according to requirements such as functional safety, resource consumption, and reaction time; the perception data will be sent to the AEB and the autonomous driving service according to business needs and processed separately; at the same time, these two services can run independently or simultaneously. The safety system detects the state information of nodes such as the autonomous driving system, and based on the priority of each control command, decides which piece of information is used as the final output.
[0104] The autonomous driving system provided by the embodiments of the present application can perform hierarchical decision-making (centralized control, edge control); it can also, based on the scene, each control node confirms the operation strategy and decision control in the corresponding scene; and the low-level nodes can quickly react according to the scene; it can realize distributed processing of calculations. When some nodes are abnormal, it can ensure that some services continue to function.
[0105] In the solution provided by the embodiments of the present application, each node evolves independently: each node evolves independently while keeping the interface unchanged.
[0106] In the autonomous driving system provided by the embodiments of the present application, the capabilities of various sensors of the autonomous driving system can be reused, and there is no need to invest additionally in an auxiliary driving system.
[0107] The autonomous driving system provided by the embodiments of the present application can reduce the complexity of the central node.
[0108] The autonomous driving system provided by the embodiments of the present application uses a hierarchical method, and sensors and low levels can be deployed nearby. The low-level processing logic is simple, the response speed is fast, it shares some capabilities of centralized control and achieves some functional safety goals.
[0109] In the autonomous driving system provided by the embodiments of the present application, each domain directly uses a unified vehicle-mounted network, which supports high-speed interconnection and the on-demand connection of each system for different vehicle models.
[0110] The autonomous driving system provided by the embodiments of the present application is based on the idea of service orientation and provides a global publishing and notification mechanism for intelligent sensors; sensors can be introduced quickly.
[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and do not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.
Claims
1. An autonomous driving system, characterized in that, Configured in a vehicle, the system includes: at least one sensor combination, each of the at least one sensor combination comprising at least one sensor; At least one edge node, wherein the edge nodes in the at least one edge node correspond to the sensor combinations in the at least one sensor combination; wherein a first edge node in the at least one edge node is configured to determine first perception information of the vehicle based on data collected by a first sensor combination corresponding to the first edge node; A central node connected to the at least one edge node, the central node is used to delegate a first state decision function to a first edge node among the at least one edge node, so that the first edge node determines that the vehicle enters the first state or the first driving strategy of the vehicle in the first state by executing the first state decision function; wherein the execution of the first state decision function requires at least the first perception information.
2. The system according to claim 1, wherein The central node is used to delegate the first state decision-making function to the first edge node under the driving state of the vehicle.
3. The system according to claim 1, wherein The central node is configured to determine the first state decision function from a plurality of preset state decision functions according to the first sensor combination.
4. The system according to claim 1, characterized in that, The system further includes at least one actuator configured to cause the vehicle to enter the first state or to execute the first driving strategy.
5. The system according to claim 4, wherein The at least one actuator is directly connected to the first edge node; the central node is used to determine the first state decision function from a plurality of preset state decision functions according to the at least one actuator.
6. The system according to claim 4, characterized in that When the first edge node determines that the vehicle needs to enter the first state or the vehicle needs to execute the first driving strategy, the first edge node sends a control instruction to the at least one actuator, so that the vehicle enters the first state or the at least one actuator executes the first driving strategy.
7. The system according to any one of claims 1-6, characterized in that, The central node is also used to send a second state decision function to a second edge node among the at least one edge node, so that the second edge node determines that the vehicle enters the second state or the second driving strategy of the vehicle in the second state by executing the second state decision function; wherein, the execution of the second state decision function requires at least second perception information, and the second perception information is determined by the second edge node based on the data collected by the sensor combination corresponding to the second edge node; the second state is different from the first state.
8. The system according to any one of claims 1-6, characterized in that, The execution of the first state decision function also requires third perception information; the third perception information is determined by a third edge node among the at least one edge node based on data collected by the sensor combination corresponding to the third edge node; the third edge node is used to send the third perception information to the first edge node when the third perception information is determined.
9. The system according to any one of claims 1-6, characterized in that, The first state is any of the following: Over the air (OTA) status, highway pilot (HWP) status, and autonomous emergency braking (AEB) status.
10. The system according to any one of claims 1-6, characterized in that, Each edge node in the at least one edge node is directly connected to a sensor in a corresponding sensor combination.
11. The system according to any one of claims 1-6, characterized in that, The edge node is an electronic control unit ECU, and the central node is a domain controller.
12. An autonomous driving system, characterized in that, Configured in a vehicle, the system includes: at least one actuator, configured to cause the vehicle to enter a first state or to execute a driving strategy in the first state; a first edge node connected to the at least one actuator; a central node connected to the first edge node, configured to delegate a first state decision function to the first edge node, wherein the first state decision function is used by the first edge node to determine whether the vehicle enters a first state or a driving strategy of the vehicle in the first state; When the vehicle needs to enter the first state or the vehicle needs to execute the first driving strategy in the first state, the first edge node sends a control instruction to at least one actuator, and the control instruction is used for the vehicle to enter the first state or enable the at least one actuator to execute the first driving strategy.
13. The system according to claim 12, wherein The system further comprises: at least one sensor; A second edge node directly connected to the at least one sensor is used to determine first perception information of the vehicle based on data collected by the at least one sensor, and send the first perception information to the first edge node, so that the first edge node performs the first state decision function based on the first perception information.
14. The system according to claim 12, wherein The first edge node is directly connected to at least one sensor; the first edge node is also used to determine the first perception information of the vehicle based on the data collected by the at least one sensor, and perform the first state decision function based on the first perception information.
15. The system according to claim 14, wherein The central node is configured to determine the first state decision function from a plurality of preset state decision functions according to the at least one sensor.
16. The system according to claim 12, wherein The at least one actuator is directly connected to the first edge node; the central node is used to determine the first state decision function from a plurality of preset state decision functions according to the at least one actuator.
17. The system according to any one of claims 12-16, characterized in that, The first edge node is an electronic control unit ECU, and the central node is a domain controller.
18. An autonomous vehicle, characterized in that, The method comprises the system according to any one of claims 1 to 11 or the system according to any one of claims 12 to 17.